In most schizophrenic patients, a prodromal phase is evident from less than a year to several years before psychosis onset, characterized by attenuated psychotic symptoms and deterioration in social, academic, and occupational function. Research on the schizophrenia prodrome has increased in recent years for at least two reasons. First, while electrophysiolgical (e.g., N100, mismatch negativity, P300, error-related negativity components of the event-related potential;ERP) and functional magnetic resonance imaging (fMRI) measures of neurocognitive dysfunction (e.g., sensory processing, attention, working memory, performance monitoring) in schizophrenia are generally assumed to predate the onset of psychosis, reflecting a genetic vulnerability and/or a neurodevelopmental insult, this critical pathophysiological hypothesis remains largely untested. Second, evidence that a longer duration of untreated psychosis after illness onset predicts a poor clinical outcome has led to the hypothesis that earlier antipsychotic treatment may improve clinical outcomes by slowing or attenuating a progressive pathophysiological process. An extension of this hypothesis, namely that antipsychotic treatment of prodromal symptoms may delay or even prevent the onset of psychosis, has already received preliminary support in two recent placebo-controlled clinical trials. However, while current clinical criteria for diagnosing the prodromal syndrome show promising predictive validity, with estimated 12-month psychosis conversion rates ranging from 20 to 60%, this still leaves 40 to 80% of patients identified as prodromal who are not at imminent risk for psychosis onset and who, therefore, may not require or benefit from early antipsychotic treatment or other interventions. Thus, enhancing the predictive validity of prodromal criteria by incorporation of neurobiological measures is a major research imperative. These considerations motivate the current study, which has three major objectives: 1) To examine whether ERP and fMRI measures of neurocognitive function, which are known to be abnormal in schizophrenia, are present in prodromal patients prior to the onset of psychosis, 2) to determine whether these abnormalities, if present, are equivalent or attenuated in severity relative to early illness schizophrenic patients, and 3) to assess whether these neurobiological abnormalities are predictive of conversion to psychosis in prodromal patients followed monthly for 24 months.